Among American men, prostate cancer is the most commonly diagnosed cancer and the second leading cause of cancer death. Experimental and clinical evidence suggests that endogenous androgens could be contributors to the etiology of prostate cancer. However, epidemiologic studies have produced ambiguous findings. Some of these studies included a relatively small number of subjects, or have measured hormone levels following the diagnosis of prostate cancer. Even among the studies that have overcome these problems, none have examined a possible interaction between hormone levels and genetic determinants either of enzymes that influence hormone metabolism or of the androgen receptor on which these hormones act. We propose to conduct a case- control study nested within the Carotene and Retinol Efficacy Trial (CARET), a randomized control trial involving 18,314 participants. We will examine blood specimens obtained from approximately 300 CARET participants who developed prostate cancer at least six months after enrollment (which began in 1985) and 300 controls matched for age, race, time of blood draw, study center, and intervention arm. We will test the hypotheses that the risk of prostate cancer is related to: 1) serum concentrations of testosterone, sex hormone binding globulin, androstenedione, dehydroepiandrosterone sulfate, 3-alpha-androstanediol glucuronide, and estradiol; 2) polymorphisms of the 5 alpha-reductase type II gene, which may influence the ability of 5 alpha-reductase to convert testosterone into dihydrostestosterone, a more active androgen; and 3) polymorphisms of the androgen receptor gene, which may influence the sensitivity of the prostate to circulating and intraprostatic hormones. Because of the size of the proposed study, the measurement of hormone levels present at least six months before diagnosis, and the assessment of genetically-determined characteristics that may influence etiology (especially when considered jointly with hormone levels), we believe the study has the potential to substantially advance our understanding of the etiology of prostate cancer.